Comment on “A Fourier transform EPR study of uracil and thymine radical anions in aqueous solution” by J. M. Lü, J. Geimer, S. Naumov and D. Beckert, Phys. Chem. Chem. Phys., 2001, 3, 952

Lü and co-workers have reported in situ photolytic experiments which produced reduction products in pyrimidine bases.¹ From the highly resolved FT EPR experiments the authors have obtained accurate hyperfine coupling constants of even the small proton couplings that are usually hidden in the linewidth of solid-state EPR spectra. In the article by Lü et al. there is discussion of several non-planar radical anions. The authors calculated spin densities and isotropic hyperfine couplings of the pyrimidine base reduction products by density functional theory to support this non-planar assignment. These conclusions are however at odds with what is observed when similar radicals are produced in the solid state.

Radicals similar to those reported by Lü et al. have been observed in irradiated single crystals and studied by detailed EPR/ENDOR experiments.² In these studies there is ample evidence that most of the radicals observed are planar. Typical results for the thymine reduction product observed in irradiated single crystals of thymine monohydrate³ are shown in Table 1. The direction of A_min (Table 1) is known to be associated with the direction of the [double bond splayed left]C–H bond, while the direction associated with the A_mid indicates the direction of the π-electron orbital. These directions are easily calculated for the crystal structure,⁴ and are included in Table 1. One sees that the direction associated with A_mid deviates only 1.1° from the computed perpendicular to the ring plane (the direction of the π-orbital), while the direction of A_min deviates only 2.5° from the computed direction of the C6–H bond. This is very clear evidence that the radical shown here is planar in the solid state.

Table 1 Hyperfine coupling parameters for pyrimidine anions

				Direction cosine
Coupling	Principal value^a	Isotropic value^a	Dipolar value^a	〈a〉	〈b〉	〈c〉
a All hyperfine couplings in MHz. b See refs. 3 and 4. c Angle that the direction of Amid makes with the perpendicular to the ring plane is 1.1°. d Angle that the direction of Amin makes with the C6–H bond direction is 2.5°. e See refs. 5 and 9. f Angle that the direction of Amid makes with the perpendicular to the ring plane is 7.1°. g Angle that the direction of Amin makes with the C6–H bond direction is 4.0°. h See refs. 6 and 10. i Angle that the direction of Amid makes with the perpendicular to the ring plane is 12.2°. j Angle that the direction of Amin makes with the C6–H bond direction is 6.1°.
Thymidine^b	−55.7		−22.7	0.732	0.656	0.185
C6–Hα	−30.3	−33.1	2.8	0.680	−0.722	−0.130^c
	−13.2		19.9	0.048	0.221	−0.974^d
5′-UMP^e	−59.1		−23.0	0.765	0.464	0.447
C6–Hα	−32.9	−36.1	3.2	−0.027	−0.669	0.742^f
	−16.2		19.9	−0.644	0.579	0.498^g
MUEA^h	−59.9		−20.8	0.263	0.953	0.149
C6–Hα	−35.3	−39.1	3.8	0.964	−0.256	−0.065^i
	−22.2		16.9	−0.024	0.160	−0.987^j

---

The experimental [double bond splayed left]C6–H_α hyperfine coupling tensors for two uracil reduction products determined from irradiated single crystals are listed in Table 1.^5,6 One notes that the agreement between experimental direction cosines and the directions of the ring perpendicular and C6–H bond directions indicate that these radical sites are essentially planar also. Lü et al.¹ report the C6–H_α experimental isotropic hyperfine coupling is −35.10 MHz for the uracil anion, which agrees well with the results obtained from detailed EPR/ENDOR experiments on a single crystal of 5′UMP (Table 1).

Lü et al.¹ report DFT calculations (B3LYP/6-31G(d)) on several reduction products. For example, the uracil anion is shown to have a fairly large C6–H_α hyperfine coupling (−48 MHz) when the C6–H proton is in the ring plane. The authors also report that for these calculations the N1–H is 14° out of the plane, while the N3–H is 11° out of the plane. All of these observations are at variance with the results reported in the solid-state. The C6–H_α proton isotropic hyperfine determined experimentally is in the range of −36 to −39 MHz, and the proton is in the molecular plane. Also these non-planar N1 and N3 protons would be easy to detect with ENDOR, but no large additional hyperfine couplings associated with N1 and N3 were observed experimentally.

The calculations performed by Lü et al.¹ involve optimizations with B3LYP and MP2 on 6-31 G(d) basis sets followed by hyperfine coupling calculations at the B3LYP/6-31 G(d) level. If one computes the hyperfine couplings for the planar uracil anion at a higher level (B3LYP/6-311+G(2df,p) on the optimized structure of 1-methyl uracil (1-MU)⁷, the C6–H_α isotropic hyperfine coupling is −40.43 MHz, which is close to the experimental values listed in Table 1 for the 1-MU anion observed in the single crystal of 1-methyl uracil : 9-ethyl adenine (MUEA).

It is important to realize that in the solid-state it is rare to observe charged radicals. Most reduction products observed in pyrimidines in the solid state are actually protonated anions.² Calculations performed on the planar [double bond splayed left]C4–OH protonated anion of 1-MU lead to a C6–H_α isotropic hyperfine coupling is −41.87 MHz, again close to the experimental value.

Calculations on both the native anion and the C4–OH protonated anion in 1-MU produce reasonable agreement with the experimental results obtained from planar radicals in the solid-state. The second entry in Table 1 is the uracil anion observed in 5′-UMP. Here the C6–H_α hyperfine coupling is −36.1 MHz, lower than in 1-MU, and more like the uracil anion reported by Lü et al.¹ (−35.1 MHz). So far no calculations on uracil anions with planar geometries have yielded C6–H_α hyperfine couplings below −40 MHz. This includes higher level optimizations and higher level spin density calculations. For example the C6–H_α hyperfine coupling using B3LYP/6-311++g(3df,3pd)//B3LYP/6-311+g(d,p) was −41.7 MHz.

Lü et al.¹ contend that the C4–C5–C6–H torsion angle must be ∼10° to reduce the C6–H_α coupling to ca. −35 MHz. However, performing these same calculations with the extended basis sets (B3LYP/6-311+G(2df,p)), it can be seen that this torsion angle needs to be less than 5° (see Table 2).

Table 2 Hyperfine coupling parameters for non-planar pyrimidine anions

Experimental coupling^a	Isotropic value^a	Torsion angle^b	Total energy^c
a All hyperfine couplings in MHz. b This angle is the C4–C5–C6–H torsion angle. c Energy units in Eh. The first entry is for the planar C2 structure. The other entries are for the non-planar Cs structures, with the last entry being the energy minimum. d The model here is 1-MU. e The model here is normal thymine.
Uracil^d	−40.42	0.0°	−454.150020
C6–Hα	−30.72	5.0°	−454.150432
−36.1	−22.40	10.0°	−454.150720
	−13.65	15.0°	−454.150924
	−5.13	19.9°	−454.150996
Thymine^e	−40.31	0.0°	−454.157367
C6–Hα	−31.06	5.0°	−454.157731
−33.1	−21.59	10.0°	−454.158108
	−11.96	15.0°	−454.158432
	−2.55	20.0°	−454.158648
	+5.23	24.3°	−454.158711

---

In solution the C6–H_α isotropic hyperfine coupling for the thymine anion is observed to be −33.00 MHz.¹ This agrees almost exactly with the results in the solid state for thymidine shown in Table 1. Theoretical calculations performed on the thymine anion are presented in Table 2. While these results are similar to those discussed above for the uracil anion, there are some significant differences. As with uracil, the planar thymine anion is not the minimum energy configuration (as previously noted by Wetmore et al.⁸), and the computed C6–H_α isotropic hyperfine coupling is too large. A search for the energy minimum produced a C6–H_α proton ca. 24° out of the plane. Unlike the case for uracil, the C6–H_α hyperfine coupling dropped to zero and then reached a positive value of about 5 MHz. This means there is a change in the orbital of the unpaired electron. As a π-radical bends, the A_iso(H_α) would change from a negative value to large positive values which are characteristic of σ-radicals in which the unpaired spin is in an orbital possessing significant s-character. Examples would be −64 MHz which is characteristic of the planar methyl radical and A_iso(H_α) = +384 MHz for the CHO σ-radical.

For the reduction product observed in thymidine in the solid state, the electron adduct is protonated at C4[double bond, length half m-dash]O.³ DFT calculations on the planar protonated thymine anion with the C4–OH proton in the molecular plane yield a C6–H_α isotropic hyperfine coupling of −40.1 MHz. As with the uracil case, it seems as if the native anion and the C4[double bond, length half m-dash]O protonated anion in thymine are indistinguishable at this level of theory.

In conclusion, it has been shown that calculations using higher order basis sets appreciably lower the C6–H_α isotropic hyperfine coupling of the thymine and uracil anions. Protonation of the thymine or uracil anions at C4[double bond, length half m-dash]O doesn't seem to affect this coupling. It appears as is a slight bending of the anion radical may be necessary to produce the experimentally observed C6–H_α hyperfine coupling. Calculations at higher level basis sets show that this out of plane bending may be less that 5°, and as such, is close to the error limits on the deviations of the direction cosines determined from the ENDOR experiments.

This work is supported by PHS Grant R01 CA36810-14 awarded by the National Cancer Institute, DHHS.

References

1. J. M. Lü, J. Geimer, S. Naumov and D. Beckert, Phys. Chem. Chem. Phys., 2001, 3, 952.
2. D. M. Close, Radiat. Res., 1993, 135, 1.
3. E. O. Hole, E. Sagstuen, W. H. Nelson and D. M. Close, J. Phys. Chem., 1991, 95, 1494.
4. D. W. Young, P. Tollin and H. R. Wilson, Acta Crystallogr., Sect. B, 1969, 25, 1423.
5. H. C. Box, W. R. Potter and E. E. Budzinski, J. Chem. Phys., 1975, 62, 3476.
6. E. Sagstuen, E. O. Hole, W. H. Nelson and D. M. Close, Radiat. Res., 1998, 149, 102.
7. M. J. Frisch et al., Gaussian 98, Gaussian, Pittsburgh, PA, 1998..
8. S. D. Wetmore, R. J. Boyd and L. A. Eriksson, J. Phys. Chem. B, 1998, 102, 5369.
9. E. Shefter and K. N. Trueblood, Acta Crystallogr., 1965, 18, 1067.
10. F. S. Mathews and A. Rich, J. Mol. Biol., 1964, 8, 89.

---